emrg: evolution_prompt — CI-check pre-LGTM gate - #645
Conversation
…any LGTM, re-trigger dropped-push CI)
argszero
left a comment
There was a problem hiding this comment.
✅ LGTM — cycle 20260810-190101. Doc-only evolution_prompt.md: CI-check pre-LGTM gate (verify gh pr checks before LGTM, workflow_dispatch re-trigger if lost) + quick-ref entry for the #644 zero-checks lesson. Verified: pytest 680 green on branch (incl. placeholder-substitution guard), template vars consistent. Sensible guard against lost push events.
argszero
left a comment
There was a problem hiding this comment.
✅ LGTM — cycle 20260810-190558. Verified the diff: doc-only change to evolution_prompt.md, template variables ({{ owner }}/{{ repo }}//) consistent with surrounding steps, no code/test impact (680). CI pass confirmed (run 31381758394). The added gate directly encodes the #644 zero-checks lesson — 'no checks reported' ≠ CI passed; re-trigger via workflow_dispatch and wait. Good hardening.
Hardening from Round 461 review of #644: the PR's push event was dropped (zero checks on branch) while both local runs were green and a parallel cycle had already posted LGTM. CI re-validation was only caught by checking
gh pr checks <N>.Adds a mandatory review step before posting any LGTM:
gh pr checks <N>— "no checks reported" means the push event was lost, NOT that CI passedgh workflow run test.yml --ref <branch>(emrg: add workflow_dispatch to Test workflow — manual CI re-trigger without empty commits #527) orscripts/re-trigger-ci.sh <branch>(emrg: add re-trigger-ci.sh — workflow_dispatch CI re-trigger, replaces empty commits #529), wait for completionAlso adds quick-ref entry. No test count change (680).